US20160380023A1

US20160380023A1 - Semiconductor optical device integrating photodiode with optical waveguide and method of forming the same

Info

Publication number: US20160380023A1
Application number: US15/193,814
Authority: US
Inventors: Yoshihiro Yoneda; Ryuji MASUYAMA; Hideki Yagi; Naoko Konishi
Original assignee: Sumitomo Electric Industries Ltd
Current assignee: Sumitomo Electric Industries Ltd
Priority date: 2015-06-29
Filing date: 2016-06-27
Publication date: 2016-12-29
Anticipated expiration: 2036-06-27
Also published as: JP2017017102A; US9893100B2; JP6421708B2

Abstract

A semiconductor optical device that integrates photodiodes (PDs) and optical waveguides coupling with the PDs and a method of forming the semiconductor optical device are disclosed. The optical waveguides in a portion in the lower cladding layer thereof provides a modified layer that forms a conduction barrier of the lower cladding layer. The modified layer is formed by converting the conduction type thereof or implanting protons therein. The modified layer prevents the electrical coupling between PDs through the waveguides.

Description

BACKGROUND

1. Field of the Invention
The present invention relates to a semiconductor optical device and a method of forming the semiconductor optical device.
2. Prior Arts
S. Farwell et al., “InP Coherent Receiver Chip with High Performance and Manufacturability for CFP2 Modules”, Optical Fiber Communication Conference 2014, USA, OSA Technical Digest, W1I.6, March 2014.
An optical device applicable to a coherent optical system provides a multi-mode interferometer (MMI) having, what is called, an optical hybrid, with a plurality of photodiodes (PDs) that convert optical signals provided from the optical hybrid into electrical signals. The PDs couple with the MMI through optical waveguides. In particular, when an MMI is integrated on a substrate with the PDs in the optical semiconductor device for the coherent optical system, conductive semiconductor layers are commonly formed in the MMI, the optical waveguides, and the PDs. The respective PDs are electrically isolated by forming trenches therebetween. That is, forming the PDs within respective mesas physically isolated from neighbor mesas by the trenches, the PDs may be electrical isolated. However, because the trenches are not extended within the optical waveguides, only existing between the PDs; electrically conduction paths are left through the conductive semiconductor layers in the optical waveguides.

SUMMARY OF THE INVENTION

An aspect of the present invention relates to a semiconductor optical device that includes a substrate, a photodiode (PD), and an optical waveguide, where the PD and the optical waveguide are integrally provided on the substrate made of semiconductor material. The PD includes first to third layers also made of semiconductor materials and stacked on the substrate in this order. The optical waveguide, which is continuous to the PD on the substrate, includes a modified layer, and fourth and fifth layers each made of semiconductor materials and stacked on the substrate. The modified layer has a thickness substantially equal to a thickness of the first layer, and the fourth layer butts against the second layer. That is, the second layer in the PD and the fourth layer in the optical waveguide form, what is called, a butt-joint. Also, the fourth layer has a bandgap wavelength shorter than a band gap wavelength of the second layer, namely, the fourth layer is substantially transparent for an optical signal subject to the PD. A feature of the semiconductor optical device of the present embodiment is that the modified layer forms a barrier for carriers conducting in the first layer.
Another aspect of the present application relates to a method of forming a semiconductor optical device that integrally includes a PD and an optical waveguide continuous to the PD on a substrate made of semiconductor material. The method includes steps of: (1) growing a first to third layers sequentially on the substrate, where the first to third layers are made of semiconductor materials, and the first and third layers have conduction types different from each other; (2) forming a modified layer in at least an area of the first layer, at which the optical waveguide is to be formed thereon, by converting the conduction type of the first layer, where the modified layer forms a pn-junction against the first layer adjacent to the modified layer; (3) forming a mesa stripe by etching the third and second layers so as to leave the first layer and the modified layer, where the mesa stripe includes the third and second layers but excludes the modified layer; (4) embedding the mesa stripe by selectively growing fourth and fifth layers sequentially on the first layer and the modified layer, the fourth and fifth layers being made of semiconductor materials different from each other, where the fourth layer butts against the second layer and has a bandgap wavelength shorter than a bandgap wavelength of the second layer; (5) forming a photodiode mesa by etching the third layer, the second layer, and a portion of the first layer, where the photodiode mesa has a height greater than a total thickness of the third layer and the second layer; and (6) forming waveguide mesa concurrently with a step of forming the photodiode mesa by etching the fifth layer, the fourth layer, and a portion of the modified layer, where the waveguide mesa having a height greater than a total thickness of the fifth layer and the fourth layer.
Still another aspect of the present application also relates to a method of forming a semiconductor optical device that integrally includes a PD and an optical waveguide continuous to the PD on a semiconductor substrate. The method includes steps of: (1) growing a first to third layers sequentially on a substrate, where the first to third layers are made of semiconductor materials; (2) forming a mesa stripe by etching the third and second layers so as to leave the first layer, where the mesa stripe includes the first to third layers; (3) forming a modified layer in an area of the first layer outside of the mesa stripe by implanting protons, where the area is adjacent to the mesa stripe and the optical waveguide is to be formed thereon, and the modified layer shows resistivity far higher than resistivity of the first layer adjacent to the modifier layer; embedding the mesa stripe by selectively growing fourth and fifth layers, each made of also semiconductor materials, sequentially on the first layer and the modified layer, where the fourth layer butts against the second layer in the mesa stripe and has a bandgap wavelength shorter than a bandgap wavelength of the second layer; (4) forming a photodiode mesa by etching the third layer, the second layer, and a portion of the first layer, where the photodiode mesa has a height greater than a total thickness of the third layer and the second layer; and (5) concurrently with a step of forming the photodiode mesa, forming a waveguide mesa by etching the fifth layer, the fourth layer, and a portion of the modified layer, where the waveguide mesa has a height greater than a total thickness of the fifth layer and the fourth layer, and the waveguide mesa is continuous to the photodiode mesa.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and other purposes, aspects and advantages will be better understood from the following detailed description of a preferred embodiment of the invention with reference to the drawings, in which:

FIG. 1 schematically illustrates a plan view of a semiconductor optical device according to the first embodiment of the present invention;

FIG. 2 shows a cross section of the semiconductor optical device taken along the line II-II indicated in FIG. 1;

FIG. 3 shows a cross section of the semiconductor optical device taken along the line III-III indicated in FIG. 1;

FIG. 4 shows a cross section of the semiconductor optical device taken along the line IV-IV indicated in FIG. 1;

FIG. 5 shows a flow chart of the process of forming the semiconductor optical device of the first embodiment shown in FIG. 1;

FIGS. 6A and 6B show the first step of forming the semiconductor optical device shown in FIG. 1;

FIGS. 7A and 7B show a step of forming the semiconductor optical device performed after forming the modified layer and embedding the mesa stripe for the photodiode mesa;

FIGS. 8A and 8B show a step of forming the modified layer in the first layer by converting the conduction type of the first layer according to the second embodiment of the present invention;

FIGS. 9A and 9B show a step of forming the semiconductor optical device performed subsequent to step shown in FIGS. 8A and 8B;

FIGS. 10A and 10B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 9A and 9B;

FIGS. 11A and 11B show a step of forming the modified layer by implanting protons in the first layer according to the third embodiment of the present invention;

FIGS. 12A and 12B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 11A and 12A;

FIGS. 13A and 13B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 12A and 12B;

FIGS. 14A and 14B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 13A and 13B;

FIGS. 15A and 15B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 14A and 14B, where steps shown in FIGS. 15A and 15B and subsequent figures are common to the second and third embodiments of the present invention;

FIGS. 16A and 16B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 15A and 15B;

FIGS. 17A and 17B show a step of forming the semiconductor optical device performed subsequent to a step shown in FIGS. 16A and 16B;

FIGS. 18A and 18B show a step of completing the process of forming the semiconductor optical device subsequent to a step shown in FIGS. 17A and 17B; and

FIG. 19 shows a structural cross section of a semiconductor optical device according to the first embodiment of the present invention.

DESCRIPTION OF EMBODIMENT

Some preferable examples of the present invention will be described as referring to drawings. In the descriptions of the drawings, numerals or symbols same with or similar to each other will refer to elements same with or similar to each other without overlapping explanations.
FIG. 1 is a plan view of a semiconductor optical device according to the first embodiment; FIG. 2 shows a cross section taken along the line II-II indicated in FIG. 1; and FIG. 3 also shows a cross section of the semiconductor optical device but taken along the line III-III indicated in FIG. 1.
Referring to FIG. 1, the semiconductor optical device 1 includes a substrate 10, photodiodes (PDs) 20, and a waveguide mesas 30. The substrate 10 provides a top surface which is divided into first to fifth areas, A₁to A₅. The first to fourth areas, A₁to A₄, are arranged along an axis AX₁, which is a longitudinal axis of the semiconductor optical device 1 and extends substantially in parallel to an optical axis of optical waveguides. The substrate 10 may be made of semi-insulating indium phosphide (InP) doped with irons (Fe).
As illustrated in FIGS. 1 and 2, the first area A₁provides the PDs 20 including four PD elements, 20 a to 20 d. The PD elements, 20 a to 20 d, are formed in respective PD mesas 20M in the first area A₁and stack first to third layers, 41 to 43, where the first layer 41 has a first conduction type, while, the third layer 43 has a second conduction type opposite to the first conduction type. The first to third layers, 41 to 43, may be made of semiconductor materials.
The waveguide mesa 30 is provided in the second to fourth areas, A₂to A₄. The waveguide mesa 30 in the second area A₂includes the first layer 41, a fourth layer 44, and a fifth layer 45 arranged on the substrate 10 in this order and each being made of semiconductor materials, where the fourth layer 44 may operate as a core layer of a waveguide, while, the fifth layer 45 may show a function of an upper cladding layer against the core layer. The fourth and fifth layers in the second area A₂form, what is called, a butt-joint against the second and third layers, 42 and 43, in the first area A₁. The third area A₃includes a modified layer 50, the fourth layer 44, and the fifth layer 45 stacking on the substrate 10 in this order. The fourth area A₄provides a stacked arrangement of the layers identical with those in the second area A₂. The fourth area A₄provides the multi-mode interference (MMI) coupler formed by the waveguides. The fifth area A₅in a portion adjacent to the first area A₁also includes the first layer 41, the fourth layer 44, and the fifth layer 45 stacked on the substrate 10 in this order. Thus, the second, fourth and fifth areas, A₂, A₄and A₅have the stacking arrangement on the substrate 10 same with each other.
As FIG. 1 illustrates, the semiconductor optical device 1 provides, in the third area A₃, the modified layers 50, which has an electrical characteristic different from that of the first layer 41 so as to interpose an electrical barrier between the first layers 41 in the second and fourth areas, A₂and A₄. The modified layer 50 may prevent the PDs 20 from electrically coupling to each other through the waveguide mesa 30. The semiconductor optical device 1 may further provide isolation mesas formed so as to surround the respective PD mesas 20M and cut the first layer 41. The isolation mesas may prevent the first layer 41 from electrically coupling to the first layer 41 in the neighbor PD elements in the lateral direction of the semiconductor substrate 10. Thus, the modified layers 50 and the isolation mesas attributed surrounding the PD mesas 20M may effectively isolate the first layer 41 from the first layer 41 in neighbor PD elements. The modified layer 50 in the third area A₃may be formed by at least one of doping impurities that convert the first layer 41 into the second conduction type and implanting protons (H⁺) therein.
Doping with Impurities Converting into Opposite Conduction Type
When the first layers 41 in the second and fourth areas, A₂and A₄, show the first conduction type, for instance, an n-type conduction; the dope of p-type impurities into the first layer 41 in the third area A₃becomes effective to convert the first layer 41 into the modified layer 50 with the p-type conduction, which forms a first pn-junction 50E against the first layer 41 in the second area A₂and a second pn-junction 50F against the first layer 41 in the fourth area A₄. Thus, an arrangement of the n-p-n structure may be formed. When the first layer 41 shows the p-type conduction, the modified layer 50 may be doped with n-type impurities to form the p-n-p structure. The n-p-n structure, or the p-n-p structure, may become a barrier for the carrier transportation, which effectively prevents the electron transportation between two n-type regions in the n-p-n structure, while, the p-n-p structure may prevent the hole transportation between two p-type regions. Thus, the first layers 41 in the second and fourth areas, A₂and A₄, may be effectively isolated by thus formed n-p-n structure, or the p-n-p structure. A combination of the isolation mesa and the modified layer 50 may realize the effective electrical isolation between PD elements, 20 a to 20 d.
The present embodiment diffuses Zn atoms, which is one of p-type dopants for InP, into an n-type InP doped with silicon (Si). The Zn concentration [Zn] in the modified semiconductor layer after the diffusion of Zn atoms becomes, for instance, 1×10¹⁷cm⁻³to 2×10¹⁸cm⁻³, which is higher than the Si concentration [Si] of, for instance, 5×10¹⁶cm⁻³to 1.5×10¹⁸cm⁻³. A diffusion process of Zn atoms may be carried out by conditions of, for instance, a di-methyl-zinc as a source gas for the Zn diffusion and a diffusion temperature of 550 to 650° C.
Implanting Protons
When the modified layer 50 is formed by implanting protons H⁺; the implanted protons H⁺induces a lot of defects within the first layer 41 to convert the layer into the modified layer 50. Implanted protons are left as interstitials between the host atoms in the modified layer 50; accordingly, the proton concentration [H⁺] in the modified layer 50 becomes higher than those in the first layers 41 in the second and fourth areas, A₂and A₄. The proton interstitials and damages caused by the implanted protons increase resistivity of the modified layer 50 compared with that of the first layers 41 in the second and fourth areas, A₂and A₄. Accordingly, the modified layer 50 may show a function of the conduction barrier between the first layers 41 of the second area A₂and that of the fourth area A₄. Thus, a coupling between the PD elements through the waveguide mesas 30 may be effectively prevented by the modified layer 50.
The present embodiment implants protons into the Si-doped n-type InP layer to modify the first layer 41, which increases the proton concentration, namely, hydrogen concentration in the modified InP layer converted from the n-type InP layer, for instance, from 1×10¹⁸to 1×10¹⁹cm⁻³. The dosage of the protons is, for instance, from 5×10¹⁴to 5×10¹⁵cm⁻², and the acceleration voltage thereof from 50 to 400 keV. The original n-type InP layer not implanted with protons has the hydrogen concentration, equivalently to the proton concentration, from 1×10¹⁷to 1×10¹⁸cm⁻³.
The waveguide mesa 30 in the structure thereof along the line III-III indicated in FIG. 1 will be described as referring to FIG. 3. The waveguide mesa 30 stacks the modified layer 50, the fourth 44, and the fifth 45 in this order in the third area A₃. The modified layer 50 in a portion thereof extends within the isolation mesa 50M. The insulating layer 22 covers the top 30A and the sides 30B of the waveguide mesa 30, and the side 50B of the isolation mesa 50M. The insulating layer 22 may be made of inorganic material containing silicon (Si), typically, silicon nitride (SiN), silicon oxide (SiO₂), silicon oxy-nitride (SiON), and so on.
The modified layer 50 wholly extends along the width 50D in the isolation mesa 50M. Also, the modified layer 50 has a thickness from the top 50A thereof to the semi-insulating substrate 10, which means that the isolation mesa 50M includes no first layer 41 which is not converted into the modified layer 50 at least in the cross section of the isolation mesa 50M. Thus, the modified layer 50 in the third area A₃may electrically isolate the first layer 41 in the second area A₂from the first layer 41 in the fourth area A₄.
The fourth layer 44, which may be made of undoped InGaAsP, extends on the modified layer 50 in the third area A₃, the first layer 41 in the second area A₂, and the first layer 41 in the fourth area A₄. Because an i-type semiconductor layer substantially shows an electrically insulating characteristic, the i-type fourth layer 44 may become an electrical barrier between the first layer 41 in the second layer A₂and the first layer 41 in the fourth area A₄in spite of these first layers 41 in the second and fourth areas, A₂and A₄, couple to each other through the i-type core layers 44 in the second to fourth areas, A₂to A₄. Thus, no electrical paths that detour the modified layer 50 may be formed between the first layers 41 in the second and fourth areas, A₂and A₄.
Practical widths, 50D and 30D, of the isolation mesa 50M and the waveguide mesa 30, respectively, in the third area A₃are shown in the table below.
bottom width 50D 2.5 to 5.0 μm

top width 30D 1.5 to 3.0 μm

Thus, the bottom width 50D, the width of the isolation mesa 50M is wider than the top width 30D, the width of the waveguide mesa 30, by about one (1) μm. Referring to FIG. 1, a distance W₁between the modified layer 50 and the edge 20E of the PD mesa 20M may be about ten (10) μm, and the modified layer 50 may have a length L₁of, for instance, 100 μm along the optical axis AX₁.
The fourth area A₄includes a waveguide structure for the MMI coupler 60, which is often called as a 90° hybrid, that provides one or more input ports and a plural output ports The MMI coupler 60 of the present embodiment includes one 2×4 MMI coupler 60 a and one 2×2 MMI coupler 60 b. The MMI coupler 60 outputs de-modulated optical signals into the waveguides, 61 a to 61 d, optically coupling the output ports of the MMI coupler 60 with the PD 20 that converts the de-modulated optical signals into electrical signals.
A table below summarizes an exemplary structure of the semiconductor optical device shown in FIGS. 1 to 3.


substrate 10	iron (Fe) doped semi-insulaling InP
first area A₁
first layer 41 (cathode)	silicon (Si) doped n-type InP
second layer 42	undoped GaInAs, or undoped GaInNAs
third layer 43 (anode)	zinc (Zn) doped p-type InP
second and fourth areas,
A₂and A₄
first layer 41	Si doped n-type InP
fourth layer 44	undoped GaInAsP with λ_BGof 1.05 μm
fifth layer 45	undoped InP
third area A₃
modified layer 50	zinc (Zn) diffused or proton (H⁺) implanted InP
fourth layer 44	undoped GaInAsP with λ_BGof 1.05 μm
fifth layer 45	undoped InP

A symbol λ_BGmeans a bandgap wavelength that corresponds to bandgap energy of a semiconductor material. As shown in the table above, the fourth layer, namely, the core layer of the optical waveguide, has the band gap wavelength of 1.05 μm shorter than that of the second layer, namely, un-doped GaInAs whose band gap wavelength is 1.55 μm or longer. That is, the fourth layer 44 is substantially transparent for an optical signal subject to the PD formed in the first area A₁. Also, the fifth layer has refractive index smaller than that of the fourth layer, which may form an optical waveguiding structure.

The PD 20 may further provide a sixth layer 46 with the second conduction type on the third layer 43. The sixth layer 46 may be made of p-type InGaAs containing p-type dopants, typically zinc (Zn). Also, the PD 20 may further provide, between the first layer 41 and the second layer 42, between the second layer 43 and the third layer 43, and/or between the third layer 43 and the six layer 46, intermediate layers that moderate a hetero barrier formed between those semiconductor layers. The intermediate layer may be made of, for instance, InGaAsP with optionally selected in respective compositions.
Referring to FIG. 4, which shows a cross section taken along the line IV-IV indicated in FIG. 1, the PD 20 provides a embedding layer 21 that covers sides of the PD mesa 20M. The embedding layer 21 may be made of iron (Fe) doped semi-insulating InP. The insulating layer 22, which may be made of inorganic material containing silicon (Si), typically, silicon nitride (SiN), silicon oxide (SiO₂), silicon oxy-nitride (SiON), and so on, covers the embedding layer 21 and the top of the PD mesa 20M as leaving an opening 23. An anode 51 p, which may be a eutectic metal of AuZn or an alloy containing platinum (Pt), is in contact to the sixth layer 46 through the opening 23. Provided on the anode 51 p is an interconnection 52 p made of gold (Au). An end of the interconnection 52 p opposite to the PD mesa 20M may provide a pad to which a wire-bonding is to be carried out.
In the first area A₁, no first layer 41 exists in a portion between the PD elements, 20 a and 20 b. That is, the PD elements, 20 a and 20 b, may be electrically isolated from neighbor PD elements. The substrate 10 in a top thereof between the PD elements is covered with the insulating layer 22 without being covered with the first layer 41. The PD elements, 20 a and 20 b, provide respective cathodes 51 n accompanying respective interconnections 52 n whose ends opposite to the cathodes 51 n provide pads to which the wire-boding is carried out. The cathode 51 n may be a eutectic metal of AuGe or alloy containing a eutectic metal, typically, AuGeNi; while, the interconnections 52 n are primarily made of gold (Au).
Next, a process of forming the semiconductor optical device will be described. FIG. 5 shows a flow chart of the process according to an embodiment of the present invention; and FIGS. 6 to 17 show respective steps of the process, where parts A in respective figures, FIGS. 6A to 17A, show cross sections taken along the line II-II indicated in FIG. 1; while, parts B in respective figures, FIGS. 6B to 17B, show cross sections taken along the lines, VIB-VIB to XVIIB-XVIIB, appearing in respective figures, FIGS. 6A to 17A.
Preparing Substrate and Growing Layers
The step S₁prepares a substrate 10 that includes first to fifth areas, A₁to A₅. The step S₂shown in FIGS. 6A and 6B grows semiconductor layers, 41 to 46, on the substrate 10; specifically, the process forms the semiconductor stack 2 by growing the first to third layers, 41 to 43, and the six layer 46 sequentially on the substrate 10 in this order by, for instance, the organic metal vapor phase epitaxy (OMVPE) technique or the molecular beam epitaxy (MBE). The third layer 43 has a conduction type different form that of the first layer 41. The second layer 42 may be an absorption layer, that is, the first to third layers, 41 to 43, may form an n-i-p structure, or a p-i-n structure for a photodiode.
The process may grow the sixth layer 46 on the third layer 43 that has the second conduction type same with that of the third layer 43. The sixth layer 46 may be made of zinc (Zn) doped indium gallium arsenide (InGaAs) to be operable as a contact layer for the anode 51 p. Also, the process may form intermediate layers between the first layer 41 and the second layer 42, between the second layer 42 and the third layer 43, and between the third layer 43 and the sixth layer 46 to moderate hetero-barriers between those layers. Thicknesses of the first, the second, and the third layers are:


	l^stlayer 41	1.5~3.0 μm
	2^nd layer 42	0.2~0.5 μm
	3^rd layer 43	1.5~3.0 μm

Forming Modified Layer
The third step S₃shown in FIGS. 7A and 7B converts the first layer 41 in the third area A₃into the modified layer 50. The conversion of the first layer 41 in the third area A₃may be carried out by one of doping dopants that converts the first conduction type of the first layer 41 into the second conduction type, or implanting protons. Processes of step S₃and subsequent thereto depend on which technique is carried out. The explanation below assumes that a step of converting the conduction type will be denoted as the step S_3-1, while, another technique of implanting protons will be denoted as the step S_3-2.
Converting Conduction Type of First Layer
The step of forming the modified layer 50 by converting the conduction type will be first described as step S_3-1. As step S_3-1shown in FIGS. 8A and 8B, the process forms a patterned mask 71 on the semiconductor stack 2, where the patterned mask 71 provides openings 72 in the third area A₃. The patterned mask 71 may be made of inorganic material, typically, silicon nitride (SiN). Then, p-type dopants are thermally diffused into the semiconductor stack 2 through the opening 72. The p-type dopants may be, for instance, zinc (Zn). Zn atoms may diffuse into the first layer 41 and reach the substrate 10, which may convert the first layer 41 into a p-type layer, as shown in FIGS. 9A and 9B. After the thermal diffusion of the p-type dopants, the patterned mask 71 is removed. When the patterned mask 71 is made of SiN, buffered hydrofluoric acid (BHF) may solve silicon nitride (SiN).
After the removal of the patterned mask 71, the process selectively etches portions of the semiconductor stack 2 to form a mesa stripe 4 as shown in FIGS. 10A and 10B. Specifically, the process forms another patterned mask 47 that extends along the direction perpendicular to the optical axis of the optical waveguide, and selectively etches the semiconductor stack 2 by, what is called, a wet-etching so as to leave the whole first layer 41, and from the second layer 42 to the sixth layer 46 in the areas out of the first area A₁. The left layers of the second and third layers, 42 and 43, and the sixth layer 46 may form the mesa stripe 4. The patterned mask 47 is to be used in a step subsequent to the selective etching of the present step. The wet-etching of the present step uses an etchant containing at least a chloric acid and an acetic acid, or another etchant containing at least a sulfuric acid and a hydrogen peroxide. The patterned mask 47 used in the present step may made of, for instance, silicon nitride (SiN).
The process for thermally diffusing the dopants may convert the first layer 41 in the conduction type thereof into the modified layer 50 having the second conduction type which is opposite to the conduction type of the original first layer 41. Thus, the modified layer 50 may form the first pn junction 50E against the first layer 41 in the second area A₂and the second pn junction 50F against the first layer 41 in the fourth area A₄. The two pn-junctions, 50E and 50F, may become a conduction barrier for carriers in the first layers 41 in the second and fourth areas, A₂and A₄, and electrically isolate the first layers 41.
Implanting Protons
Next, another process S_3-2of implanting protons will be described as referring to FIGS. 11A and 11B. The process S_3-2first forms the mesa stripe 4 in advance to the implantation by steps same with those aforementioned step S_3-1. The mesa stripe 4, which is formed in the first area A₁, includes the second and third layers, 42 and 43, and the sixth layer 46 on the first layer 41.
Then, as shown in FIGS. 12A and 12B, a patterned mask 73 covers the mesa stripe 4 and a whole surface of the first layer 41 out of the third area A₃. That is, the patterned mask 73 provides openings 74 in the third area A₃. The patterned mask 73 may be formed by the photolithography for a photoresist and the opening 74 may be a rectangular or a square. Subsequently, the protons H⁺ are implanted into the first layer 41 in the third area A₃through the opening 74 to form the modified layer 50. The protons H⁺ reach the substrate 10, that is, the range of the implantation exceeds the thickness of the first layer 41, as shown in FIGS. 13A and 13B; accordingly, the first layer 41 in a whole of the third area A₃may be converted into the modified layer 50. Typical conditions of the proton implantation are:


acceleration voltage	80 to 200	keV
temperature	350 to 400°	C.
dosage
1 × 10¹⁵to 3 × 10¹⁵	cm⁻²

The modified layer 50 which is implanted by the protons H⁺ inherently shows high resistivity because the implanted protons H⁺ becomes interstitials and/or induces defects in the first layer 41, which degrades the crystal quality and considerably increases resistivity of the first layer 41 compared with that in the first layer 41 in the second area A₂and in the fourth area A₄. A layer with considerably high resistivity may be a conduction barrier and electrically isolate the first layers 41 in the second and fourth areas, A₂and A₄. Thus, the proton implantation into the first layer 41 in the third area A₃may form the modified layer 50 therein.
Thus, step S_3-1of converting the conduction type and/or step S_3-2of implanting protons may form the modified layer 50 and the mesa stripe 4.
The process then performs re-growth of semiconductor layers, 44 and 45, without removing the patterned mask 47 for forming the mesa stripe 4 so as to form the semiconductor stack containing the fourth and fifth layers, 44 and 45, in the second to fifth areas, A₂to A₅, as shown in FIG. 14A. The fourth and fifth layers, 44 and 45, in the second to fifth areas, A₂to A₅, which are un-doped layers, are provided for the core layer 44 and the upper cladding layer 45, respectively. The fourth layer 44 butts against the second layer 42 in the first area A₁and may have a thickness substantially equal to a thickness of the second layer 42. Also, the fifth layer 45 butts against the third layer 43 and the sixth layer 36 in the first area A₁and may have a thickness substantially equal to a total thickness of the third layer 43 and the sixth layer 46. In an alternative, the forth layer 44 may be thicker than the second layer, for instance about 10% thicker than the second layer; while, the fifth layer 45 may have a thickness compensating the thickened forth layer 44 such that the top surface thereof is leveled with the top surface of the sixth layer 46. Thus, the fourth and fifth layers, 44 and 45, in the second area A₂forms the butt-joint against the second to sixth layers, 42 to 46, in the first area A₁. Accordingly, the mesa stripe 4 is often called as the butt-joint mesa. The patterned mask 47, which is prepared to form the mesa stripe 4 and to embed the mesa stripe 4, is removed after the re-growth of the layers, 44 and 45, by a buffered hydrofluoric acid.

Forming PD Mesa and Waveguide Mesa

Step S₄shown in FIGS. 15A and 15B patterns another mask 75 on the semiconductor stacks in the first to fifth areas, A₁to A₅, where the patterned mask 75 determines a plane shape of the PD mesa 20M and a planar shape of the waveguide mesa 30. The patterned mask 75 may be made of silicon nitride (SiN). The process etches the fifth, fourth and a portion of the first layers or the modified layer, 45, 44, and 41 or 50, in the second to fifth areas, A₂to A₅, also etches the sixth, third, second, and a portion of the first layers, 46, 43, 42, and 41, between the PD elements, 20 a to 20 d, in the first area A₁. Thus, the fourth layer 44, the fifth layer 45, and the first layer 41 in the second to fifth areas, A₂to A₅, may operate as the core layer, the upper cladding layer, and the lower cladding layer in the waveguide mesa 30; while, the second layer 42, the third layer 43, the sixth layer 46, and the first layer 41 in the PD mesa 20M may operate as the absorption layer, the anode layer, the contact layer, and the cathode layer for the respective PD elements, 20 a to 20 d.
Passivating PD Mesa
Then, the process forms still another patterned mask 77 without removing the aforementioned mask 75 in step S₅as shown in FIGS. 16A and 16B. The patterned mask 77, which may have a thickness of about 100 nm and may be made of silicon nitride (SiN), has an opening within which the PD mesa 20M is exposed, but covers the sides of the waveguide mesa 30 and a most surface of the first layer 41 in the fifth area A₅except around the PD mesa 20M. A portion of the waveguide mesa 30 in the second area A₂continuous to the PD mesa 20 is exposed in the sides thereof within the opening. Then, a Fe-doped InP 21 is selectively grown on the side of the PD mesa 20M and the surface of the first layer 41 exposed within the opening of the patterned mask 77; accordingly, the Fe-doped InP 21 reliably passivates the sides of the PD mesa 20M, which may reduce leak currents flowing from the third layer 43 (the anode layer) to the first layer 41 (the cathode layer) along the sides of the PD mesa 20M.
Forming Isolation Mesa
Step S₆shown in FIGS. 17A and 17B removes the first layer 41 and a part of the substrate 10 around the PD mesa 20M and the waveguide mesa 30. Specifically, removing the patterned mask 75, another mask 76 made of silicon nitride (SiN) is patterned so as to cover the PD mesa 20M and the waveguide mesa 30. An etching may remove the first layer 41 around the PD mesa 20M and a part of the substrate 10 around the PD mesa 20M and the waveguide mesa 30 in the first to fifth areas, A₁to A₅. Thus, because the first layer 41 left between the PD mesas 20M and between the waveguide mesas 30 may be removed; the isolated PD mesas 20M and the isolated waveguide mesas 30 may be obtained on the substrate 10.
Forming Insulating Layer and Electrodes
Step S₇shown in FIGS. 18A and 18B forms an insulating layer 22, and anode and cathode electrodes, 51 p and 51 n, respectively. The insulating layer 22, which protects the PD mesa 20M and the waveguide mesa 30, may be made of inorganic material containing silicon (Si), typically, silicon nitride (SiN), silicon oxide (SiO₂), silicon oxy-nitride (SiON), and so on. The present embodiment provides the insulating layer 22 made of SiN deposited by the plasma-enhanced chemical vapor deposition (p-CVD) technique. The insulating layer 22 has a thickness of, for instance, 200 nm. Removing the insulating layer 22 partially so as to expose the top of the PD mesa 20M by the opening 23 and the first layer 41 in the side of the PD mesa 20M, the anode 51 p and the cathode 51 n are formed within the respective openings so as to be in contact to the six layer 46 in the top of the PD mesa 20M and the first layer 41 in the side of the PD mesa 20M, respectively. The anode 51 p may be made of eutectic alloy of AuZn, or alloy containing platinum (Pt); while, the cathode 51 n may be made of eutectic alloy of AuGe, or AuGe containing nickel (Ni). After the formation of the anode 51 p and the cathode 51 n, interconnections, 52 p and 52 n, for respective electrodes, 51 p and 51 n, are carried out. Thus, the process to form the semiconductor optical device 1 is completed.
FIG. 19 schematically illustrates a stacking arrangement of the semiconductor layers of the present invention. FIG. 19 assumes that the first layer 41 has the n-type conduction. The following table summarizes a typical configuration of the respective layers.


substrate 10	Fe-doped semi-insulating InP
first layer 41	Si-doped InP operable as a cathode and a lower
	cladding layer
modified layer 50	p-InP doped with p-type dopants, or InP implanted
	with protons
fourth layer 44	undoped InGaAsP operable as a core layer
fifth layer 45	undoped InP operable as an upper cladding layer
second layer 42	undoped InGaAs operable as an absorption layer
third layer 43	Zn-doped p-type InP operable as an anode

When the first layer 41 shows the p-type conduction, the configuration of the respective layers, 41 to 45, may be exchanged between the n-type conduction and the p-type conduction. Also, the first layer 41 may be exchanged from Si-doped layer to Zn-doped layer, the second layer 42 may be exchanged from Zn-doped to Si-doped.
While particular embodiments of the present invention have been described herein for purposes of illustration, many modifications and changes will become apparent to those skilled in the art. Accordingly, the appended claims are intended to encompass all such modifications and changes as fall within the true spirit and scope of this invention.

Claims

We claim:

1. A semiconductor optical device, comprising:

a substrate made of semiconductor material;

a photodiode (PD) including first layer, a second layer, and a third layer sequentially stacked on the substrate, where the first layer, the second layer, and the third layer are made of semiconductor materials; and

an optical waveguide continuous to the PD, the optical waveguide including a modified layer, a fourth layer, and a fifth layer on the substrate, the modified layer having a thickness substantially equal to a thickness of the first layer, where the fourth layer butts against the second layer and has a bandgap wavelength shorter than a bandgap wavelength of the second layer,

wherein the modified layer forms a barrier for carriers conducting in the first layer.

2. The semiconductor optical device of claim 1,

wherein the modified layer has a conduction type different from a conductor type of the first semiconductor layer and forms a pn-junction against the first semiconductor layer.

3. The semiconductor optical device of claim 2,

wherein the modified layer has resistivity far greater than resistivity of the first semiconductor layer.

4. The semiconductor optical device of claim 3,

wherein the modified layer is a semi-insulating layer.

5. The semiconductor optical device of claim 1,

wherein the PD forms a photodiode mesa including the first semiconductor layer, the second semiconductor layer, and the third semiconductor layer, the PD mesa in sides thereof being passivated with a semiconductor layer.

6. The semiconductor optical device of claim 1,

wherein the PD forms a photodiode mesa including the first semiconductor layer, the second semiconductor layer, and the third semiconductor layer, and the optical waveguide forms a waveguide mesa continuous to one of sides of the PD mesa, the waveguide mesa including the modified layer, the fourth semiconductor layer, and the fifth semiconductor layer, and

wherein the substrate has hollows around the PD mesa and the waveguide mesa.

7. The semiconductor optical device of claim 1,

further including a multi-mode interference (MMI) coupler on the substrate,

wherein the optical waveguide optically couples the MMI coupler with the PD.

8. A method of forming a semiconductor optical device that integrally includes a photodiode (PD) and an optical waveguide on a substrate made of semiconductor material, optical waveguide having an optical axis, the method comprising steps of:

growing a first layer, a second layer, and a third layer sequentially on a substrate, where the first layer, the second layer and third layers are made of semiconductor materials, and the first layer and the third layer have respective conduction types different from each other;

forming a modified layer in at least an area of the first layer, at which the optical waveguide is to be formed thereon, by converting the conduction type of the first layer, the modified layer forming a pn-junction against the first layer adjacent to the modified layer;

forming a mesa stripe by etching the third layer and the second layer so as to leave the first layer, where the mesa stripe extends along a direction perpendicular to the optical axis of the optical waveguide and includes the first layer, the second layer, and the third layer, but excludes the modified layer;

embedding the mesa stripe by selectively growing a fourth layer and a fifth layer sequentially on the first layer and the modified layer, where the fourth layer and the fifth layer are made of semiconductor materials different from each other, and the fourth layer butts against the second layer in the mesa stripe and has a bandgap wavelength shorter than a bandgap wavelength of the second layer;

forming a photodiode mesa by etching the third semiconductor layer, the second semiconductor layer, and a portion of the first semiconductor layer, where the photodiode mesa has a height greater than a total thickness of the third layer and the second layer; and

forming waveguide mesa concurrently with a step of forming the photodiode mesa by etching the fifth layer, the fourth layer, and a portion of the modified layer, where the waveguide mesa has a height greater than a total thickness of the fifth layer and the fourth layer.

9. The method of claim 8,

wherein the step of forming the modified layer is carried out by diffusing dopants thermally, the dopants converting the conduction type of the first layer.

10. The method of claim 8,

wherein the modified layer forms a pn-junction against the first semiconductor layer adjacent to the modified layer.

11. The method of claim 8,

further including a step of, after the step of forming a photodiode mesa and the current step of forming the waveguide mesa, covering sides of the photodiode mesa by a semiconductor layer as exposing sides of the waveguide mesa.

12. The method of claim 8,

further including a step of, after the step of forming a photodiode mesa and the concurrent step of forming the waveguide mesa, partially etching the substrate in exteriors of the photodiode mesa and the waveguide mesa.

13. A method of forming a semiconductor optical device that includes a photodiode (PD) and an optical waveguide integrally on a substrate made of semiconductor material, the optical waveguide having an optical axis, the method comprising steps of:

growing a first layer, a second layer, and a third layer sequentially on a substrate, the first layer, the second layer and the third layer being made of semiconductor materials;

forming a mesa stripe by etching the third layer and the second layer so as to leave the first semiconductor layer, where the mesa stripe includes the first layer, the second layer, and the third layer, and extends along a direction perpendicular to the optical axis of the optical waveguide;

forming a modified layer in an area of the first layer by implanting protons, where the area is adjacent to the mesa stripe and the optical waveguide is to be formed thereon, and the modifier layer shows resistivity far greater than resistivity of the first layer adjacent to the modified layer;

embedding the mesa stripe by selectively growing a fourth layer and a fifth layer sequentially on the first layer and the modified layer, the fourth layer and the fifth layer being made of semiconductor materials, where the fourth layer butts against the second layer in the mesa stripe and has a bandgap wavelength shorter than a bandgap wavelength of the second layer;

forming a photodiode mesa by etching the third layer, the second layer, and a portion of the first layer, the photodiode mesa having a height greater than a total thickness of the third layer and the second layer; and

concurrently with the step of forming the photodiode mesa, forming a waveguide mesa by etching the fifth layer, the fourth layer, and a portion of the modified layer, where the waveguide mesa has a height greater than a total thickness of the fifth layer and the fourth layer, the waveguide mesa is continuous to the photodiode mesa.

14. The method of claim 13,

further including a step of, after the step of forming the photodiode mesa and the concurrent step of forming the waveguide mesa, covering sides of the photodiode mesa by a semiconductor layer as exposing sides of the waveguide mesa.

15. The method of claim 13,

further including a step of, after the step of forming the photodiode mesa and the concurrent step of forming the waveguide mesa, partially etching the substrate in exteriors of the photodiode mesa and the waveguide mesa.